光子学报

• 图像与信号处理 • 上一篇    下一篇

基于结构相似度的图像融合质量评价

狄红卫;刘显峰   

  1. (暨南大学光电工程研究所,广州 510632)

  • 收稿日期:2006-01-04 修回日期:2006-03-04 出版日期:2006-05-25 发布日期:2006-05-25
  • 通讯作者: 狄红卫 E-mail:tdhw@jnu.edu.cn

Image Fusion Quality Assessment Based on Structural Similarity

Di Hongwei;Liu Xianfeng

  

  1. (Institute of Optoelectronic Engineering, Jinan University, Guangzhou 510632)
  • Received:2006-01-04 Revised:2006-03-04 Online:2006-05-25 Published:2006-05-25
  • Contact: Di Hongwei E-mail:tdhw@jnu.edu.cn

摘要: 在分析现有图像融合质量评价方法特点的基础上,提出了新型的、基于结构相似度的图像融合质量评价方法.针对不同情况,分别采用平均结构相似度、加权平均结构相似度、结构信息与交互信息量之乘积作为图像融合质量客观评价标准.该方法充分考虑了图像的结构信息和人类视觉系统的特性,可以为不同场合下选择不同的算法提供依据.对不同融合算法的质量评价结果表明,该方法是一种有效的图像融合质量评价方法.

关键词: 图像融合, 结构相似度, 图像质量评价, 人眼视觉系统

Abstract: How to assess the image quality is one of the most important problems in image fusion. By analyzing the existent assessments of image fusion quality, a novel assessment of image fusion quality based on structural similarity is presented. According to different cases, average structural similarity, weighted average structural similarity, product of structural similarity and mutual information are introduced to assess image fusion quality as the objective methods. This assessment considers the structural information of image and the characteristics of human visual system, and can be used to choose the best algorithm in image fusion applications. The experiments of different kinds of image fusion algorithms demonstrate that it is an effective assessment of image fusion quality.

Key words: Image fusion, Structural similarity, Image quality assessment, Human visual system

中图分类号: